Unbabel’s TowerLLM AI model outperforms OpenAI’s GPT-4 in translation #AItranslation

Unbabel says new TowerLLM AI model beats OpenAI's GPT-4 at translation

Unbabel, a tech company offering machine and human-based translation services, has developed a new AI model called TowerLLM that outperforms OpenAI’s GPT-4o and other AI systems in translating between English and six commonly spoken languages. The model was tested against various competitors and showed superior accuracy, particularly in professional domains like finance, medicine, law, and technical writing. TowerLLM’s success suggests that newer AI systems may challenge GPT-4’s dominance in translation tasks. The model was trained on multilingual text and fine-tuned with high-quality translations, resulting in its superior performance. Unbabel plans to expand TowerLLM’s language support and enhance its capabilities for specific translation tasks, such as transcreation to capture cultural nuances. The model is available in two sizes and supports 18 language pairs. While TowerLLM excels in translation, GPT-4 may still have advantages in other tasks like reasoning, coding, writing, and summarization. Unbabel aims to continue improving TowerLLM’s performance and expanding its language support to meet the diverse needs of businesses.

Source link

Source link:

What do you think?

Leave a Reply

GIPHY App Key not set. Please check settings

Gigafactory of Compute, Zoom Meetings x AI Clones, WhatsApp AI Features | Daily AI News

Daily AI News: Gigafactory, Zoom AI Clones, WhatsApp Features #AIRevolution

OSF HealthCare research contributes to FDA approval of AI sepsis tool

OSF HealthCare research aids FDA approval of AI sepsis tool #healthcareinnovation